Introduction

Jim Kelly once said, “There’s not one person who’s even been through our training camp who could cover him. Nobody could cover him one-on-one”. Jim Kelly wasn’t referring to Andre Reed (HoF WR) or James Lofton, he was referring to Steve Tasker who only had 51 career receptions over 14 years in the NFL. Steve Tasker was 5’9’’ was arguably the best gunner in NFL history making 7 Pro Bowls and was the only special teams player to ever earn Pro Bowl MVP (1993). A gunner is a position on the special teams who’s sole role is to break free from defenders during a punt to tackle the returner and limit their return yardage. The goal of my analysis is to build on the initial research done by Michael Lopez to add specific metrics to evaluate Gunners in the NFL. (https://operations.nfl.com/gameday/analytics/stats-articles/visualizing-the-special-teams-gunner/ )

We will be using 2020 tracking and scouting information to create two models to evaluate Gunner effectiveness during punts. The models are:
1. Tackle Opportunity Probability Model
2. Expected Gunner Distance at Punt Reception over Expected

Tackle Probability

The play below is 3 yard punt return where Nsimba Webster (#14) and David Long (#25) are vying to either limit return yardage by way of tackle opportunity creation or force a fair catch. Webster here is credited with the tackle.

If we break down the tackle opportunity by frame, we see that Webster had the higher probability of tackle up given his burst past the Vise (#29) but dropped in probability when #18 on the Punt return came to help to shield. We think that although Webster was credited with the tackle, he may have gotten a bit lucky with the returner running back into him. There’s a few things we can learn from this play, which goes into how to evaluate Gunner performance.

Visualizing all of Webster’s routes we see his angle of pursuit can be a contributing factor for his Tackle Opportunities (Green representing tackles & Blue representing missed tackles).

Expected Gunner Distance at Punt Reception Model

As we saw in the Tackle Opportunity Probability Model, the distance from the ball at first major event was one of the key features in determining probability of an expected tackle opportunity. As we can see in the distribution chart below, the distribution of Gunner’s distance for each punt varies by a lot, which can be a key area for Gunner’s to separate themselves in terms of improving their tackle opportunities or forcing fair catch.

In looking at punts that were returned, we can see a small relationship of the Gunner’s ability to limit return yardage if they are within 10 yards of the punt returner (wide variability in relating the two metrics to each other).

This leads us to the creation of the “expected distance to ball” model, which was trained using a Random Forest. This helps us to understand, which players are best at angle of pursuit. It also helps us visualize, which players did better than expected at first major event (i.e. Punt Reception or Fair Catch). The purpose of this model is to show, which players were better than expected in the pursuit, thus limiting return yardage and forcing fair catch. In our sample play below, the dashed grey line will show the expected ball distance for Nsima Webster (which resulted in a tackle). We can see that at the point of a ball kick, we get a prediction for his expected distance.

To do: -Add endzone logos for each team -Add feature model for Tackle Probability -Show a summary chart or stats on performance of 2020 season for top Gunners -Which plays have the best odds at the punt and why?

-Create a model for expected return yardage?

show Nsima Webster distribution of expected ball distance

Show increase in probability due to shift in variables

Best in offset situations